Reading Between the Lines in Autism Spectrum Disorder

Lead Research Organisation: University of Bristol
Department Name: Education

Abstract

During everyday communication people read between the lines. For example, if somebody was to approach you in the street and ask "do you know where the post-office is?" based on your knowledge, you would assume they are not only asking if you know the post office's location (in which case, a simple answer of "Yes" may suffice), but also that you give them directions to the post-office, even though they did not directly ask you this. This is also true during reading; we frequently make assumptions and read between the lines to aid our understanding of a text. For example, if reading 'Sally forgot her umbrella. Her hair was soaked when she arrived at work' you might assume that it had rained and this had caused Sally's hair to be wet, even though this is not explicitly stated in the text. Reading between the lines is more formally known as inference generation and is something that people seemingly do automatically during reading, to extract unwritten, implicit information from language. Inference generation is critical for effective reading comprehension, communication, and social interactions.

Autism Spectrum Disorder (ASD) is a developmental condition characterised by restricted and repetitive patterns of behaviour (e.g., a strong aversion to change, special interests), difficulties with social interaction and communication (e.g., difficulties with eye contact), and a range of information processing differences (e.g., sensory sensitivities). Autistic individuals are also vulnerable to reading comprehension challenges and this appears to be related to inference generation skill. For example, autistic people are often reported to have difficulties answering questions about a text that requires an inference to be generated, and no current theory of reading or autism can explain this difficulty. Educational practice in relation to literacy instruction is directly informed by scientific understanding of the cognitive (mental) processes that underlie reading. To make current education techniques more effective and appropriate for autistic individuals (e.g., text design or learning tools), and to assure autistic individuals can exploit written methods of communication (e.g., text, email), it is vital to develop scientific understanding of why inference generation is challenging for autistic people.

This project will do exactly this - we will use eye tracking, arguably the best methodology to study reading processes, to examine why inference generation may be challenging for autistic adults. As a person reads, they move their eyes approximately 4 times a second, making saccades and fixations. Saccades are fast eyeball rotations that relocate the point of focus from word to word. Fixations are periods between saccades when the eyes remain relatively still for visual information encoding. The duration of fixations and patterns of saccades provide detailed information about how text is processed. We will compare eye movement patterns of autistic and non-autistic readers as they process texts that depict scenarios where inferences must be formed for successful comprehension. We will identify cognitive processing differences causative to inference generation difficulties in ASD.

This work will significantly increase scientific understanding of reading in ASD, with the long-term goal of ultimately informing the development of supportive educational and remedial practices, and contributing to the improvement of social inclusion, quality of life, and independence for autistic people. This will be achieved by communicating this work with the scientific community in Journal articles and presentations at national and international conferences. We will also communicate our research findings to stakeholders and the public via online media (website and Twitter) and collaborative events. These events will deliver community perspectives and interdisciplinary insights for future translational steps with respect to the project findings.

Publications

10 25 50